As data rates increase to meet demand for higher data throughput, signals transmitted through a communication channel are increasing susceptible to corruption by frequency-dependent signal loss of the communication channel, such as inter-symbol interference (ISI), and other noise, such as crosstalk, echo, signal dispersion, and distortion. A decision feedback equalizer (DFE) may be used to remove ISI and other noise by using a feedback loop based on previously decided symbols from the received signal. However, a conventional DFE is limited to performing the ISI removal for a symbol in a single symbol period (also referred to as a “unit interval” or UI), which is the baud rate of the communication channel. This timing constraint may limit the operations of the DFE.
Accordingly, it would be desirable and useful to provide an improved way of implementing decision-feedback equalization that facilitates meeting a growing demand for higher throughput.